Filters








2,195 Hits in 5.7 sec

A Latent Transformer for Disentangled Face Editing in Images and Videos [article]

Xu Yao, Alasdair Newson, Yann Gousseau, Pierre Hellier
2021 arXiv   pre-print
To tackle these limitations, we propose to edit facial attributes via the latent space of a StyleGAN generator, by training a dedicated latent transformation network and incorporating explicit disentanglement  ...  Our model achieves a disentangled, controllable, and identity-preserving facial attribute editing, even in the challenging case of real (i.e., non-synthetic) images and videos.  ...  Consequently, a face editing method should rely on disentangled attributes and permit identity-preserving manipulations.  ... 
arXiv:2106.11895v2 fatcat:czi7onsp75e3bm4zontcks7or4

Disentangled Lifespan Face Synthesis [article]

Sen He, Wentong Liao, Michael Ying Yang, Yi-Zhe Song, Bodo Rosenhahn, Tao Xiang
2021 arXiv   pre-print
The generated face image given a target age code is expected to be age-sensitive reflected by bio-plausible transformations of shape and texture, while being identity preserving.  ...  In this work, a novel LFS model is proposed to disentangle the key face characteristics including shape, texture and identity so that the unique shape and texture age transformations can be modeled effectively  ...  In this work, for the first time we propose a LFS model that explicitly disentangles a learned latent face representation into shape, texture and identity.  ... 
arXiv:2108.02874v2 fatcat:djxn4yaysvf7zb2lulryptydhy

FaceShapeGene: A Disentangled Shape Representation for Flexible Face Image Editing [article]

Sen-Zhe Xu, Hao-Zhi Huang, Shi-Min Hu, Wei Liu
2019 arXiv   pre-print
The conditional label-to-face transformer, which is trained in an unsupervised cyclic manner, performs part-wise face editing while preserving the original identity of the subject.  ...  On the basis of the FaceShapeGene, a novel part-wise face image editing system is developed, which contains a shape-remix network and a conditional label-to-face transformer.  ...  The transformer F learns to transform a whole-face label map to a photo-realistic face image, while preserving the identity of a conditional input image.  ... 
arXiv:1905.01920v1 fatcat:xo64iczferfrdjccxlrbcx32ea

Face Identity Disentanglement via Latent Space Mapping [article]

Yotam Nitzan, Amit Bermano, Yangyan Li, Daniel Cohen-Or
2020 arXiv   pre-print
Specifically, disentangled latent representations allow generative models to control and compose the disentangled factors in the synthesis process.  ...  By learning to map into its latent space, we leverage both its state-of-the-art quality, and its rich and expressive latent space, without the burden of training it.  ...  ACKNOWLEDGMENTS We thank the anonymous reviewers for their comments. We also thank Tal Hassner, Yuval Nirkin, Aviv Gabbay and Mingchao Sun for their help and useful suggestions.  ... 
arXiv:2005.07728v3 fatcat:3jbzg6peq5h2hgo7p67sgivmre

Only a Matter of Style: Age Transformation Using a Style-Based Regression Model [article]

Yuval Alaluf, Or Patashnik, Daniel Cohen-Or
2021 arXiv   pre-print
preserving the input identity.  ...  Finally, we demonstrate that the end-to-end nature of our approach, coupled with the rich semantic latent space of StyleGAN, allows for further editing of the generated images.  ...  ACKNOWLEDGMENTS We would like to thank Elad Richardson, Kfir Goldberg, Ohad Fried, Yotam Nitzan, and Zongze Wu for their fruitful discussions and early feedback.  ... 
arXiv:2102.02754v2 fatcat:3ut6yy66kjcyhcbc7rauw2cgxq

Semantic and Geometric Unfolding of StyleGAN Latent Space [article]

Mustafa Shukor, Xu Yao, Bharath Bhushan Damodaran, Pierre Hellier
2021 arXiv   pre-print
We thus propose a new method to learn a proxy latent representation using normalizing flows to remedy these limitations, and show that this leads to a more efficient space for face image editing.  ...  Generative adversarial networks (GANs) have proven to be surprisingly efficient for image editing by inverting and manipulating the latent code corresponding to a natural image.  ...  Identity preservation is enforced by minimizing the loss between the features extracted from a pretrained face recognition model F before and after editing.  ... 
arXiv:2107.04481v1 fatcat:i4ryd4hokrfj5kfkkfewiebfhq

Enjoy Your Editing: Controllable GANs for Image Editing via Latent Space Navigation [article]

Peiye Zhuang, Oluwasanmi Koyejo, Alexander G. Schwing
2021 arXiv   pre-print
Our model permits better control for both single- and multiple-attribute editing while preserving image identity and realism during transformation.  ...  Classic approaches for this task use a Generative Adversarial Net (GAN) to learn a latent space and suitable latent-space transformations.  ...  We study two types of transformations T : i) global and ii) local. A global transformation T refers to a semantic latent-space transformation identical for all z during inference.  ... 
arXiv:2102.01187v3 fatcat:jsprls7hcjcv7bu6i33vszpbia

FaceController: Controllable Attribute Editing for Face in the Wild [article]

Zhiliang Xu, Xiyu Yu, Zhibin Hong, Zhen Zhu, Junyu Han, Jingtuo Liu, Errui Ding, Xiang Bai
2021 arXiv   pre-print
Face attribute editing aims to generate faces with one or multiple desired face attributes manipulated while other details are preserved.  ...  By simply employing some existing and easy-obtainable prior information, our method can control, transfer, and edit diverse attributes of faces in the wild.  ...  Acknowledgments This work was supported by the National Program for Support of Top-notch Young Professionals and the Program for HUST Academic Frontier Youth Team 2017QYTD08, which are given to Dr.  ... 
arXiv:2102.11464v1 fatcat:x2y46expq5dqdb4rwhw6kxsrcu

StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery [article]

Or Patashnik, Zongze Wu, Eli Shechtman, Daniel Cohen-Or, Dani Lischinski
2021 arXiv   pre-print
Next, we describe a latent mapper that infers a text-guided latent manipulation step for a given input image, allowing faster and more stable text-based manipulation.  ...  and real images.  ...  Finally, for edits that require identity preservation, we use the identity loss defined in eq. (2).  ... 
arXiv:2103.17249v1 fatcat:hlh4h3t35vffhaisg4c2c6p5au

Third Time's the Charm? Image and Video Editing with StyleGAN3 [article]

Yuval Alaluf, Or Patashnik, Zongze Wu, Asif Zamir, Eli Shechtman, Dani Lischinski, Daniel Cohen-Or
2022 arXiv   pre-print
Finally, we introduce a novel video inversion and editing workflow that leverages the capabilities of a fine-tuned StyleGAN3 generator to reduce texture sticking and expand the field of view of the edited  ...  benefits of using the StyleSpace for fine-grained editing.  ...  Further, observe that the face identity is well-preserved for unrelated edits and that local edits, such as those changing hairstyle and expression, do not alter unrelated image regions (e.g., expression  ... 
arXiv:2201.13433v1 fatcat:ztxf6x32lvb6rearrlefmkrqge

LEED: Label-Free Expression Editing via Disentanglement [article]

Rongliang Wu, Shijian Lu
2020 arXiv   pre-print
The idea is to disentangle the identity and expression of a facial image in the expression manifold, where the neutral face captures the identity attribute and the displacement between the neutral image  ...  Two novel losses are designed for optimal expression disentanglement and consistent synthesis, including a mutual expression information loss that aims to extract pure expression-related features and a  ...  Two novel losses are designed for optimal identity-expression disentanglement and identity-preserving expression editing in training the proposed method.  ... 
arXiv:2007.08971v1 fatcat:wtbaxrdenndxxgahoe7tsn75ay

State-of-the-Art in the Architecture, Methods and Applications of StyleGAN [article]

Amit H. Bermano and Rinon Gal and Yuval Alaluf and Ron Mokady and Yotam Nitzan and Omer Tov and Or Patashnik and Daniel Cohen-Or
2022 arXiv   pre-print
It aims to be of use for both newcomers, who wish to get a grasp of the field, and for more experienced readers that might benefit from seeing current research trends and existing tools laid out.  ...  Among StyleGAN's most interesting aspects is its learned latent space. Despite being learned with no supervision, it is surprisingly well-behaved and remarkably disentangled.  ...  Examples of prominent works leveraging latent space embedding. (a) Nitzan et al. [2020] disentangle identity from other face attributes and recompose them to generate novel face images.  ... 
arXiv:2202.14020v1 fatcat:qu3plbdnszdujcwxwq3zizysje

Scaling-up Disentanglement for Image Translation [article]

Aviv Gabbay, Yedid Hoshen
2021 arXiv   pre-print
In this work, we propose OverLORD, a single framework for disentangling labeled and unlabeled attributes as well as synthesizing high-fidelity images, which is composed of two stages; (i) Disentanglement  ...  : Learning disentangled representations with latent optimization.  ...  We demonstrate in Fig. 6 that while the baselines rely on a pretrained face identity loss [7] we are able to preserve the identity better.  ... 
arXiv:2103.14017v2 fatcat:r3sinwhijbadvamkpvq7rndgay

Disentangled and Controllable Face Image Generation via 3D Imitative-Contrastive Learning [article]

Yu Deng, Jiaolong Yang, Dong Chen, Fang Wen, Xin Tong
2020 arXiv   pre-print
We propose DiscoFaceGAN, an approach for face image generation of virtual people with disentangled, precisely-controllable latent representations for identity of non-existing people, expression, pose,  ...  Experiments show that through our imitative-contrastive learning, the factor variations are very well disentangled and the properties of a generated face can be precisely controlled.  ...  embed real images into the latent space and edit the factors in a disentangled manner.  ... 
arXiv:2004.11660v2 fatcat:zaj3gkgyczdb3by7kvpcwjb3ne

StyleFlow: Attribute-conditioned Exploration of StyleGAN-Generated Images using Conditional Continuous Normalizing Flows [article]

Rameen Abdal, Peihao Zhu, Niloy Mitra, Peter Wonka
2020 arXiv   pre-print
We evaluate our method using the face and the car latent space of StyleGAN, and demonstrate fine-grained disentangled edits along various attributes on both real photographs and StyleGAN generated images  ...  For example, for faces, we vary camera pose, illumination variation, expression, facial hair, gender, and age.  ...  Face identity score: To measure the quality of the edit and quantify the identity preservation of the edits, we evaluate the edited images using a face identity score.  ... 
arXiv:2008.02401v2 fatcat:wbzz4t23uvcozk3nav675heewm
« Previous Showing results 1 — 15 out of 2,195 results